Print | Rate this content

CUSTOMER ADVISORY: HP EVA XCS10000000 reserves more spare capacity than previous firmware versions

SUPPORT COMMUNICATION - CUSTOMER ADVISORY

Document ID: c02931184

Version: 1

CUSTOMER ADVISORY: HP EVA XCS10000000 reserves more spare capacity than previous firmware versions
NOTICE: The information in this document, including products and software versions, is current as of the Release Date. This document is subject to change without notice.

Release Date: 2011-07-22

Last Updated: 2011-07-22


DESCRIPTION

Disk drive failure protection reservation (sparing) is designed to ensure that even if you have allocated 100% disk space to LUNs, Controller Software (XCS) will ensure that a single (or double) drive loss will be completely handled and will allow a return to a fully redundant state without user intervention. Sparing (or reserving) space in a disk group is shared among all volumes in the Disk Group, regardless of the number of redundant parity structures in that disk group. The more physical disks (volumes) that are included in the disk group, the less the capacity needs to be reserved on each drive. As a result, the larger the disk group, the less you might notice this change.

How sparing is implemented

When a user requests a disk drive failure protection level, controller software will always look at the worst case scenario based on the RAID types that can be used in a disk group and redundant parity structure configuration within the Disk Group. To do this, VRaid1 space is considered, and then separately, the space required for R0, R5 and R6 parity options is considered. Controller Software also calculates how much usable space is left remaining based on RAID type.

For example, if a single drive fails, the mirror pair of that drive can no longer be used for VRaid1 space, as the requested redundancy cannot be provided. As a result, Controller Software calculates that the largest drives (and their mirrors) can no longer be used. It is assumed that all of the data that can be used in this disk group is going to be R1. This results in space reservation based on the largest volume in the disk group.

If a drive failure occurs and this results in a redundant parity structure having an insufficient number of drives available for allocation, a redundant parity structure merge operation will have to complete prior to the sparing operation completing—but there will be adequate space (with the new calculations) to complete both operations—although additional time for the operations will be required.

Controller Software must also take into consideration how the volume mirror pairs are matched up. For example, if a 146 GB drive is paired with a 450 Gb drive, the maximum space available for R1 on that pair will be the lesser of the 2 volumes (146GB). The revised calculations will ignore the remaining space on the larger drive for the purposes of R1 individual drive and total disk group capacity calculations.

Based on the configuration of the redundant parity structures, a worst case scenario can occur in small disk groups and redundant parity structures that will result in considerable capacity being reserved by the disk drive failure protection level selected.

After Controller Software has calculated the required space to reserve for both mirror and parity options, the larger of the two values will be reserved.

In previous Controller Software versions, Controller Software did not take into consideration the capacity lost with the failed drive that was being assumed to contribute to the needed sparing capacity. In cases where customers have used 99% to 100% capacity prior to the failure, instances have occurred where not enough capacity was reserved to allow the sparing operation to complete. As a result, the drive was in a hung and degraded (non-redundant) state until additional space was added, or a virtual disk was removed, and free space was made available.

SCOPE

The disk drive failure protection level capacity reservation change was released in XCS10000000. With Controller Software upgrades from previous XCS versions on EVA 4400/6400/8400, you might see significant available capacity changes for a smaller disk group.

RESOLUTION

The change made in XCS 10000000 includes the same calculations for the amount of space that needs to be reserved, but it changes how much of that space is to be reserved on each drive. The total disk group capacity in the following calculation is changed to not include the space from the largest drive in the disk group that may fail. A quick calculation will note that this results in a significant change for small systems.

The space on each drive reserved for disk drive failure protection level (S) is: 2LC * r.

Where:

  • 2 is the constant for mirror pairs

  • L is the number of spares requested

  • C is the Capacity of the current drive

  • r is the "critical ratio" of current drive, derived by r = Total Drive Capacity / Diminished Disk

  • Group Capacity

NOTE: Total Drive Capacity = number of disks in disk group * capacity.

Diminished Disk Group Capacity = Total Drive Capacity minus the capacity of the number of drives associated with the selected spare policy, zero, two, or four).

The following examples consist of round numbers; but in reality, these numbers are not always round numbers.

Example of single disk drive failure protection level with a disk group of eight 1 TB disks

For VRaid1, Controller Software calculates a need for the capacity of two disk drives becoming unavailable with a single drive failure. This will result in the following calculation:

2 * 1 * 1 TB * (1 TB / 6 TB) = 333 GB reserved on each drive. This now totals 2.667 TB of reserved space in the disk group (.667 TB more space reserved as with the previous Controller Software version)

Example of dual disk drive failure protection level with a disk group of twelve 450 GB disks:

For VRaid1, Controller Software calculates a need for the capacity of four disk drives becoming unavailable with a double drive failure. This will result in the following calculation:

2 * 2 * 450 GB * (450 GB / 3600 GB) = 225 GB reserved on each drive. This totals 2700 GB of reserved space in the disk group (900 GB more space reserved as with the previous Controller Software version).

Workaround

To lessen the effects of this change, you can perform the following after upgrading your firmware to XCS10000000:

  1. Change the disk group protection level to none or single (if it was set to double ) before adding more drives to disk group.

  2. Delete some unused virtual disks from the disk group if the array still does not have enough space to create new virtual disks after option #1 above, and more drives cannot be added in a timely manner.

  3. Migrate some virtual disks to a different disk group that has more available disk space.

Proactive Updates

Receive support alerts (such as Customer Advisories), driver updates, software, firmware, and customer replaceable components, in your e-mail through HP Subscriber's Choice. Sign up for Subscriber's Choice Driver, Patch, Security, and Support alerts at the following URL: http://www.hp.com/go/myadvisory


Hardware Platforms Affected: HP 4400 Enterprise Virtual Array, HP 6400/8400 Enterprise Virtual Array, HP EVA P6000 Storage
Components Affected: Not Applicable
Operating Systems Affected: Not Applicable
Software Affected: Not Applicable
Third Party Products Affected: Not Applicable
Support Communication Cross Reference ID: IA02931184
©Copyright 2011 Hewlett-Packard Development Company, L.P.
Hewlett-Packard Company shall not be liable for technical or editorial errors or omissions contained herein. The information provided is provided "as is" without warranty of any kind. To the extent permitted by law, neither HP or its affiliates, subcontractors or suppliers will be liable for incidental,special or consequential damages including downtime cost; lost profits;damages relating to the procurement of substitute products or services; or damages for loss of data, or software restoration. The information in this document is subject to change without notice. Hewlett-Packard Company and the names of Hewlett-Packard products referenced herein are trademarks of Hewlett-Packard Company in the United States and other countries. Other product and company names mentioned herein may be trademarks of their respective owners.

Provide feedback

Please rate the information on this page to help us improve our content. Thank you!